Accelerating offshore renewable energy deployment through AI models of the ocean

The design of offshore renewable energy systems should consider realistic ocean extremes which can be complex and highly nonlinear. However, linear models are often used for design due to their low cost, resulting in uncertainty. This project will develop AI models for nonlinear water wave problems, primarily aiming to learn the spatio-temporal mapping from linear (easy to model, widely used) to fully nonlinear wave fields.

Both fully nonlinear potential flow models (e.g. OceanWave3D), and smoothed particle hydrodynamics (SPH) models that capture wave breaking, will be used to train the model, covering a wide range of realistic extreme conditions.

The outcome will be an open-source model which will give fast yet accurate fully nonlinear extreme kinematics based on a simplified linear input, which can subsequently be used to drive fast models for offshore system design. Findings comes at a critical time for the offshore renewable energy sector as we look to accelerate the design and deployment of floating offshore wind turbines globally.